3183
2008, 112, 3183-3186 Published on Web 02/09/2008
Tuning the Wetting Properties of Multiwalled Carbon Nanotubes by Surface Functionalization Bhalchandra A. Kakade and Vijayamohanan K. Pillai* Physical & Materials Chemistry DiVision, National Chemical Laboratory, Pune 411008, India ReceiVed: December 11, 2007; In Final Form: January 28, 2008
A simple strategy of chemical functionalization is employed here for controlling the wettability of multiwalled carbon nanotube films from superhydorphobic (156°) to nearly hydrophilic (40°), which further shows superhydrophilicity with a water contact angle of about 0° within 2 s. The hydrophobic surface (water contact angle 145°) in case of nitric-acid-treated MWCNT bucky paper shows an unusual pinning action even at a tilt angle of 45°, following Wenzel behavior of the surface. The variation in the wetting properties of MWCNT bucky paper using such surface designing strategies provides controlled heterogeneity on the surface with no or minimal effect on surface roughness.
Variation of both the chemical composition and the geometrical structure of solid surfaces1,2 to control the wetting behavior is of enormous importance for many processes in living organisms3-5 and for numerous industrial applications such as design of self-cleaning eyeglass, windows, and adhesive fixtures. As a result, many superhydrophobic surfaces have been formed by a wide variety of methods ranging from lithographic fabrication to the transformation of a simple plastic by surface functionalization6-8 as well as by surface roughening.9-16 For example, a temperature-induced fine control over wettability of colloidal-crystal films of latex spheres has been reported, where surface roughness is tuned by the temperature effect.17 Similarly, a fluoroalkylsilane-coated aligned CNT forest has recently shown super-amphiphobicity with water and oil due to chemical interactions pertaining to the fluoroalkylsilyl groups.2 Because carbon nanotubes have a perfect graphitic network of sp2 carbon, they exhibit intrinsic hydrophillic behavior (CA < 90°), where further functionalization or textured arrangement could facilitate easy control of their wetting properties. Although many chemical functionalization treatments have been carried recently out on both single and multiwalled CNTs, no reports are available to date on their functionalizationdependent tuning of wetting behavior despite the importance. Alternatively, surface properties of CNTs, especially if they are in the form of paper or mats, offer the possibility of using them in more compact form for easy processing. Finally, the interactions of CNTs (both pristine as well as functionalized) with various liquids are also important for understanding a variety of new phenomena including voltage generation up on liquid flow for developing flow sensors.18 The wetting properties of single-shell CNTs have effectively been shown to be crucial for understanding capillary action and the transition from wetting to nonwetting using various liquids, which could, in principle, facilitate to do the solution chemistry inside the open ends of * Corresponding author. E-mail:
[email protected]. Fax: +91 20 25902636. Tel: +91 20 25902588.
10.1021/jp711657f CCC: $40.75
nanotubes.19-21 Indeed, the wetting properties of fibers may differ significantly from those of plane solid surfaces and the effect of curvature (diameter dependence) can be important.22 For example, Mattia et al. have shown a disordered wetting of CVD grown carbon films and nanopipes using both polar and nonpolar liquids.23 All of the above results demonstrate the significance of tuning the surface properties of CNTs for various applications. For example, a combination of surface roughness/inhomogeneity (geometrical) and surface energy (chemical) can make CNT surfaces superhydrophobic if they are aligned vertically with a diameter of few micrometers.24 Tuning of hydrophobicity/ hydrophilicity of CNT surfaces is one of the daunting challenges in order to engineer them into proper geometry and device applications due to a lower value of surface tension (∼27 mN/ m). Unfortunately, no one has demonstrated the ability to control the hydrophobicity/hydrophilicity of MWCNTs in the form of bucky paper, on the basis of chemical functionalization. Hence, a correlation between the surface functionalization and externally manifested wetting characteristics of a surface has tremendous utility in order to understand a variety of phenomena including anti-fogging, self-cleaning ability, and adaptability toward synthesis of novel and robust hybrid materials. Here we report for the first time a method of tuning the hydrophobicity of CNTs in the form of bucky paper using unique functionalization strategies. The extent of the hydrophobicity/ hydrophilicity of the bucky paper can be tuned remarkably by surface functionalization using different routes. Surface roughness is engineered through covalent coupling, facilitating a change in both the surface chemical composition and the geometrical microstructure to generate hierarchical structures where the water contact angle (CA) can be anywhere between 40° to 156°. Our results indicate that this type of controlling the hydrophobic surfaces of bucky papers on the side walls of MWCNTs could be useful for many potential applications because of the low surface energy with tunable contact angle. © 2008 American Chemical Society
3184 J. Phys. Chem. C, Vol. 112, No. 9, 2008
Letters
TABLE 1: Functionalization Procedures for Each Sample along with Dominant Functional Groups. sample
treatment
A B C D
as synthesized sample ozonolyzed MWCNTs refluxed in 3:1 nitric acid for 8 h refluxing and soaking in 3:1 nitric acid refluxed in acid mixture (1:1; 78% HNO3 and 98% H2SO4) covalent coupling with tridecylamine (TDA) microwave treatment in 1:1; 78% HNO3 and 98% H2SO4 for 4 min covalent coupling with 1,8-diaminooctane (ODA)
E F G H
functional group graphitic C (sp2) -COOCH3 -OH (major), -COOH -OH (major), -COOH -OH, -COOH -CONH(CH2)12-CH3 -OH, -COO- (major), -SO3H -CONH(CH2)8-CONH2
In the first experimental approach, the surface of MWCNTs has been modified using different oxidizing treatments such as ozonolysis (B), refluxing in 3:1 nitric acid (C), refluxing in 3:1 nitric acid followed by soaking in the same acid (D), refluxing in acid mixture (1:1; 78% HNO3 and 98% H2SO4; E), microwave treatment in presence of acid mixture (G), and covalent coupling with aliphatic mono (tridecylamine; F) and diamine (1,8-diaminooctane; H) to create the diverse functionality. Experimental details have been given in the Supporting Information (summarized in Table 1). In the second approach, the surface of MWCNTs is modified by varying the molecularlevel corrugation associated with the lesser amount of functional groups created by treatment with nitric acid. The pristine sample has been designated as “A”. Herein, we have used less crystalline CVD nanotubes (with varying diameter) for our measurements. But, CNTs with more uniform diameter and good crystallinity (especially Arc-derived) would have given more reliable results. The modified samples of carbon nanotube were characterized using diffused reflectance Fourier transform infrared (DRIFT), sessile drop water contact angle (CA) measurements, scanning electron microscopy (SEM), energy dispersive analysis of X-rays (EDX), X-ray photoelectron spectroscopy (XPS), and Raman spectroscopy. CA measurements have been performed at room temperature (27 °C) and constant humidity. The superimposed DRIFT spectrum (Figure 1) of both pristine and modified MWCNT samples gives clear evidence of the surface modification. For example, in the high-frequency region, the broad band at 3730 cm-1 corresponding to the free -O-H bonds is invariant with surface modification, although a slight shift (at 3636 cm-1) in the case of sample G is observed, which could be attributed to intermolecular bonding (H-bonding) among various functional groups on side walls. The fundamental bands corresponding to CdC (1525 cm-1), CdO (broad 1860 cm-1), and -OH (phenolic; 1250 cm-1) are also observed except in case of sample G, where peaks due to CdO and -OH are absent and a new peak at 1600 cm-1 appears, attributed to -COO- groups. In addition, bands at 2864 and 2921 cm-1, respectively, due to C-H stretching vibrational modes produced during reaction strongly indicate the presence of sp3 defects after functionalization of the MWCNTs surface. Although no quantitative relationship between the degree of functionalization and wettability has been developed, the relationship is surely a direct one, where a systematic variation in the CA has been recorded with a standard deviation of (2°. This effect of tuning the wettability of bucky papers could also be worth mentioning by the relative intensities of O1s peak in the XP spectra (Supporting Information S7).
Figure 1. Superimposed DRIFT spectrum of pristine MWCNTs (A) and surface functionalized bucky paper samples (B-H), revealing the changes in surface functional groups. Sample G shows an additional peak at 1600 cm-1 indicating the formation of -COO- groups.
Figure 2. (a) Histogram exhibiting the variations in the contact angles ((2°) of the pristine MWCNT sample (A) and modified bucky paper samples (B-H), indicating the effect of surface functionalization on the wetting behavior of MWCNT papers (b) Schematic of surface modification after chemical treatment.
As shown in the histogram in Figure 2a, an advancing contact angle of the pristine MWCNT sample is 156°, which undergoes a systematic variation with respect to different functionalization protocols. It is well known that poor wettability of the pristine CNT sample makes the drop bounce back from the surface, making it difficult to detach a drop from the tip of the pipette. The value of the water contact angle on the pristine sample is reasonably smaller as compared to that on densely packed aligned CNTs (166°) because of random contacts between solid and liquid affecting the contour, length, and continuity of the
Letters
Figure 3. Water-drop profile on sample C at different tilt angles, revealing a surface pinning action due to a lesser extent of surface functionalization, following Wenzel’s model.
triple contact line. It has also been observed that the tilt angle over the pristine sample is 16°, which is rather high compared to that for an ideal superhydrophobic surface (e3°). This higher value of the tilt angle could be attributed to the residual water content on the surface, which is clearly evidenced in the FTIR spectra. However, it clearly demonstrates that the presence and extent of different functional groups on the surface of CNTs determine the surface topography and consequently the hydrophobicity/hydrophilicity (as demonstrated in the schematic view of the bucky paper surface in Figure 2b). It means that hierarchical structures created by subnanolevel manipulation of surface topology give a novel approach to constructing promising materials for various applications. Figure 3 indicates the water-drop profile on sample C (nitric acid treated) at different tilt angles in order to monitor the sliding angle effects caused due to nano- and microstructures. Surprisingly, a very high rolling (tilt) angle of more than 45° (the water droplet sticks to the surface with variations in its shape) is observed with contact angle of 145°, revealing a surface pinning action due to a lesser extent (not yet quantified) of surface functional groups, which bind the droplet without allowing it to spread or evaporate easily. The effect of droplet pinning along with a hydrophobic surface is an indication of the Wenzel formalism, where it is assumed that the liquid fills up the space between the protrusions on the surface.25 Interestingly, this superhydrophobic film (pristine sample) shows an advancing contact angle for water of about 156° and a roll-off angle for a 4 µL droplet of about 16° (see Supporting Information S2). It has also been seen that the initial water CA of 156° on the pristine sample of a droplet volume of 4 µL evaporates within 30 min following constant contact angle (with a very small decrease in CA) until the weight of the drop decreases to an appreciable value and then follows the constant area mode below CA < 90°. After nitric acid treatment, the advancing contact angle for water becomes only 145°, with a roll-off angle of more than 45° indicating the role of functional moieties present on the surface rather than surface roughness. Surprisingly, sample G shows an initial water contact angle of 40° but spreads (0°) within 2 s, indicating the formation of a “superhydrophilic” bucky paper surface. Such a superhydrophilic surface can be well understood by 2D26,27 or 3D28,29 capillary effects on the hydrophilic surfaces. It means that a microwave treatment for 4 min can make the CNT surface superhydrophilic, where the formation of more side-wall defects in the form of various groups such as -COOH, -OH, and -SO3H (∼30 wt % oxygenated groups as judged from XPS studies) plays an important role in surface wettability. More interestingly, the sample (see in Supporting Information S4) prepared by treatment with 3:1 H2SO4 shows an initial CA value of 145°, whereas it diminishes to 0° within 6 min indicating surface pinning followed by absorption through capillary action. In sharp contrast to the pioneering work by Birdi and Vu,30,31 it has been found that the initial water CA of 145° (hydrophobic) diminishes to 0° (hydrophilic) in a linear
J. Phys. Chem. C, Vol. 112, No. 9, 2008 3185 fashion but follows the constant area mode. These variations could be due to fundamental drawbacks in their studies such as the use of different substrates and varying liquid drops. In the present case, we have adopted several functionalization strategies to tune the surface topography of the MWCNTs, which also leads to decapping (breaking of the fullerene caps) during treatment in all cases. However, the presence of different amounts of functional groups on the side walls as well as in the interior of the nanotubes leads to the variation in the wetting properties. Furthermore, when open nanotubes contact a liquid surface the internal wetting force due to the liquid surface tension acting on the nanotube interior surface should be considered equally. In the present case, sample G (microwave treated in the presence of acid mixture) shows a very small CA because of the higher extent of internal wetting along with external wetting due to the presence of a large amount of functional groups. However, sample C (nitric acid treated) still shows superhydrophobic behavior, although a rolling angle of more than 45° has been observed indicating a strong pinning action. It means that a lower extent of functional groups on the side walls (as well as on the interior of the nanotubes) makes the droplet reside on the surface but resist capillary action through the nanotube. Recently, Barber et al. have shown an enhanced internal wetting for the open-ended individual MWCNTs using the Wilhelmy force balance method.32 They have also explained that the water exhibits a significantly larger interaction with the nanotube (as a function of nanotube diameter) as compared to other organic liquids such as polyethylene glycol and glycerol. However, in our case the CA measurements have been performed on a MWCNT film comprising nanotubes with nonuniform diameter distribution, which could give an ensemble effect on wetting. Because the defect sites increase with increase in the diameter,22 the wetting properties of CVD grown nanotubes (more strained walls) will be governed by the number of defects at the nanotube surface. In particular, in the case of pristine sample, the wide diameter distribution (20-100 nm) leads to the superhydrophobic behavior. Alternatively, in the case of sample G the CNTs with larger diameter, containing a high degree of surface defects, would give smaller contact angles when compared to those of thinner nanotubes. In short, the uneven distribution of functional groups along the side walls of individual CNTs may also play a considerable role in the contact angle measurements. Consequently, the interplay of all of these factors on the wetting behavior of bucky paper prevents the quantification of contributions from capillarity and internal wetting. From the above points, it could be concluded that the variation in the wetting properties of MWCNT bucky paper using such surface designing strategies provides controlled heterogeneity on the surface with no or minimal effect on surface roughness. The achievement of tuning the wetting characteristics of MWCNT surfaces implies that the fabrication of CNT composites with an amazing range of superhydrophobic to superhydrophilic properties could be realized by improving the surface functionalization strategies. Practical utilization of such tunable CNT surfaces to make high-performance polymer composites (e.g., controlling the interfacial adhesion between CNT surface and polymer) would significantly open up new perspectives in the preparation of various polymer composites extending the range of their possible applications. Such capabilities would allow one to construct much more sophisticated architectures from MWCNTs, including electronic interconnects, robust nanocomposites, and structured thin films.
3186 J. Phys. Chem. C, Vol. 112, No. 9, 2008 Acknowledgment. We thank A. B. Gaikwad for SEM studies. University Grants Commission (UGC), Government of India and CSIR (through NMITLI program) are acknowledged for the financial support. We also thank Dr. Sivaram, Director NCL, for valuable discussion. Supporting Information Available: Experimental procedure, actual CA profiles, SEM, EDS, XPS, and Raman data of all MWCNT films are provided in detail. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Blossey, R. Nat. Mater. 2003, 2, 301-306. (2) Feng, L.; Li, S.; Li, Y.; Li, H.; Zhang, L.; Zhai, J.; Song, Y.; Liu, B.; Jiang, L.; Zhu, D. AdV. Mater. (Weinheim, Ger.) 2002, 14, 1857-1860. (3) Barthlott, W.; Neinhuis, C. Planta 1997, 202, 1-8. (4) Gao, X.; Jiang, L. Nature 2004, 432, 36-36. (5) Feng, L.; Zhang, Z.; Mai, Z.; Ma, Y.; Liu, B.; Jiang, L.; Zhu, D. Angew. Chem., Int. Ed. 2004, 43, 2012-2014. (6) McHale, G.; Schirtcliffe, N. J.; Aqil, S.; Perry, C. C.; Newton, M. I. Phys. ReV. Lett. 2004, 93, 036102-036104. (7) Oner, D.; McCarthy, T. J. Langmuir 2000, 16, 7777-7782. (8) Erbil, H. Y.; Demirel, A. L.; Avci, Y.; Mert, O. Science 2003, 299, 1377-1380. (9) Odom, T. W.; Huang, J.-L.; Kim, P.; Lieber, C.M. J. Phys. Chem. B 2000, 104, 2794-2809. (10) Chen, W.; Fadeev, A. Y.; Heieh, M. C.; O ¨ ner, D.; Youngblood, J.; McCarthy, T. J. Langmuir 1999, 15, 3395-3399. (11) Wu, Y.; Sugimura, H.; Inoue, Y.; Takai, O. Chem. Vap. Deposition 2002, 8, 47-50. (12) Tsujii, K.; Yamamoto, T.; Onda, T.; Shibuchi, S. Angew. Chem., Int. Ed. Engl. 1997, 36, 1011-1012.
Letters (13) Tadanaga, K.; Katata, N.; Minami, T. J. Am. Ceram. Soc. 1997, 80, 3213-3216. (14) Nakajima, A.; Abe, K.; Hashimoto, K.; Watanabe, T. Thin Solid Films 2000, 376, 140-143. (15) Bico, J.; Marzolin, C.; Que´re´, D. Europhys. Lett. 1999, 47, 220226. (16) Xiao, W.; Huang, Z.; He, Z. Appl. Phys. Lett. 2006, 89, 083101083103. (17) Wang, J.; Wen, Y.; Hu, J.; Song, Y.; Jiang, L. AdV. Funct. Mater. 2007, 17, 219-225. (18) Ghosh, S.; Sood, A. K.; Kumar, N. Science 2003, 299, 1042-1044. (19) Dujardin, E.; Ebbesen, T. W.; Hiura, H.; Tanigaki, K. Science 1994, 265, 1850-1852. (20) Dujardin, E.; Ebbesen, T. W.; Krishnan, A.; Treacy, M. M. J. AdV. Mater. 1998, 10, 1472-1475. (21) Ebbesen, T. W. J. Phys. Chem. Solids 1996, 57, 951-955. (22) Neimark, A. V. J. Adhes. Sci. Technol. 1999, 13, 1137-1154. (23) Mattia, D.; Bau, H. H.; Gogotsi, Y. Langmuir 2006, 22, 17891794. (24) Liu, H.; Zhai, J.; Jiang, L. Soft Matter 2006, 2, 811-821. (25) Wenzel, R. N. Ind. Eng. Chem. 1936, 40, 988-994. (26) Wang, R.; Hashimoto, K.; Fujishima, A.; Chikuni, M.; Kojima, E.; Kitamura, A.; Shimohigoshi, M.; Watanabe, T. Nature 1997, 388, 431432. (27) Sakai, N.; Wang, R.; Fujishima, A.; Watanabe, T.; Hashimoto, K. Langmuir 1999, 14, 5918-5920. (28) Harminghaus, S. Europhys. Lett. 2000, 52, 165-170. (29) Martines, E.; Seunarine, K.; Morgan, H.; Gadegaard, N.; Wilkinson, C. D. W.; Riehle, M. O. Nano. Lett. 2005, 5, 2097-2103. (30) Birdi, K. S.; Vu, D. T. J. Phys. Chem. 1989, 93, 3702-3703. (31) Birdi, K. S.; Vu, D. T. J. Adhes. Sci. Technol. 1993, 7, 485-493. (32) Barber, A. H.; Cohen, S. R.; Wagner, H. D. Phys. ReV. B 2005, 71, 115443/1-115443/5.
